Subtle subtilis Strains
How strain typing helped us nail where this organism, ubiquitous in manufacturing sites, was causing problems. Part of our MicroPhyles series.
Bacillus subtilis was in everything! At least, that’s what our test results said. The test plates for raw materials, bulk product formulations, and products filled on different lines were all covered in B. subtilis. All of these samples were tested the same day. A couple days later, all the results from a different test session looked the same.
What was going on? The above scenario, which played out at a lab I used to work for, quickly became a conundrum for those of us involved in the investigation.
We were confident the contamination came from the lab. But this wasn’t going to help get the product out the door. We needed to know (and remediate) a more specific source. To make matters worse, a third test session yielded similar results, adding more to the restricted inventory. Hundreds of thousands of dollars’ worth of inventory was sitting in a warehouse when it should have been selling on store shelves.
How could we pinpoint the exact source? Bacillus subtilis is one of the most common microorganisms found in regulated manufacturing sites. It is so common, those sites need to prove they can recover it from their products if it happens to get into them (Per USP <61> and USP <71> ). It’s one of only six species called out for these tests!
In a non-sterile consumer goods plant, the microbiologists have some of the best awareness of cleanliness and sanitization. Unfortunately, the nature of the lab offers some of the best opportunities for microbes to grow. The variety of media available is like an all-you-can-eat buffet. Incubators offer organisms a warm and cozy environment. Even the requirements from USP <61> and USP <71> mean a stock of B. subtilis will always be available, if not in use, in the lab.
To start the lab investigation, we swabbed a bunch of areas of interest. The incubators, countertops, supply cabinets, the media water bath, autoclave supplies, the test hood, pipettes and a few other spots. Any guesses where we saw B. subtilis?
Almost everywhere. All low counts. Nothing terrifying in any particular spot. The lab was designed as non-sterile, doubling as office space for five microbiologists. Removing B. subtilis from all these locations would have been an excessive undertaking. Extra hours of sporicidal wipe-downs could have been added to the daily cleaning regime to remove it and prevent it from coming back, but no one wants to be responsible for that. Besides, the lab cleaning practices that were already in place had clearly been good enough to prevent issues like this in the past.
To focus our investigation, we needed to know which spot was causing our problem. How could we do that?
This was the perfect scenario where knowing an identification beyond species level could aid the investigation. We had strain typing performed on each of our B. subtilis swab recoveries. We compared the strains to the strain that kept showing up in our product tests. There was one location where the strain exactly matched the product test results . . . the inside of the autoclave glove!
After getting this critical piece of the puzzle, the picture looked like this:
· Test media was autoclaved in glass bottles to ensure sterility
· The bottle caps weren’t fully sealed to prevent explosions during autoclaving
· Some of the media bubbled up during the heating process, allowing some residue to get on the exterior of the bottle near the cap.
· The microbiologist donned the autoclave glove to transfer media bottles from the autoclave tray to a water bath. (The bath ensures the media remains a liquid for pour plate testing; at the time, media could be used up to 24 hours after autoclaving)
· When the bottles cooled, the microbiologist tightened the caps. At this point, it was likely that the microbiologist transferred the B. subtilis from the interior cloth liner of the glove to the media residue on the outside of the bottle.
· When the bottles were poured, the surface tension of the media caused it to contact the areas near the rim of the bottle that were contaminated, resulting in a series of contaminated plates.
Thanks to the strain typing test, we immediately knew the failing results were not representative of the product*. We tested the retain samples - back-up samples collected along with the original test samples for investigations like this. Everything looked great. We released the lots and freed up our backlog.
The lab was ecstatic that their cleaning regimen didn’t need an overhaul. We started wiping media bottles with IPA after placement into the water bath. We also purchased a new autoclavable autoclave glove. The new glove didn’t have the cloth liner that harbored microbes. That was it. We never saw a scenario like that again.
*Strain typing helped with another aspect of our investigation. Even though we had other strains of B. subtilis in the lab, we didn’t see them in our test samples. The results of this investigation showed strains in these locations were not contaminating the test session, i.e. the current cleaning practices and aseptic techniques were generally successful at preventing contamination from these locations. Since we focused our investigation on the exact source with strain typing, we avoided increasing lab cleaning frequencies and adding new cleaning locations (like the inside of cabinets). Those extra cleanings would have been a huge resource drain while providing little value.
This is part of our MicroPhyles series, an ongoing feature detailing interesting case studies from the world of microbial taxonomy.